Error due to Recording/Setting parameters

[RokasSt]

Hi Richard,

Would be helpful if you can look at this. I specified the following combination of parameters:
in the RecordingSettings array:
v_m: 1
maxRecTime: 100
sampleRate: 20
In the SimulationSettings array:
simulationTime: 1000
timeStep: 50
parallelSim: 0

and I get the error at the end of running the model:
Attempted to access RS.dataWriteSteps(11); index out of bounds because numel(RS.dataWriteSteps)=10.

Error in simulate (line 158)
if simStep == RS.dataWriteSteps(numSaves)

Error in runSimulation (line 128)
simulate(TP, NP, SS, RS, NeuronIDMap, NeuronModelArr, ...

Also, when the simulation time in ms is set to 200 000 (200 s), I also get the same error: attempted to access... index out of bounds .....

Best Regards, Rokas
@richardtomsett

[richardtomsett]

Hi @RokasSt hope things are going well!
I guess you should also get the following warning at the start of the simulation:

Incompatible timeStep supplied: 50
Setting timeStep to 32 ms
Sample rate incompatible with timeStep
Sampling instead at 31 Hz

?
So the code that detects and manages incompatible time steps or sample rates assumes that your time step is always going to be <= 1, which is why it doesn't fix the time step you specify properly. Is there a reason you're using such a big time step of 50 ms (and such a slow sample rate)?

[RokasSt]

Hi,

Thanks for prompt reply. The reason behind is that some cell models
in lems/neuroml need simulation times larger than sec and time steps, in some
examples, are in tens of ms ( as here 50 ms). However, these cell
models are rather old and newer versions behave more naturally
in ms scale. I might actually work with them in due course, so
I am not asking you to change anything in the master, I just
wanted to know the reason.

Thanks!
Regards, Rokas
@richardtomsett

---

From: Richard [notifications@github.com]
Sent: 03 September 2015 17:25
To: OpenSourceBrain/VERTEXShowcase
Cc: Stanislovas, Rokas
Subject: Re: [VERTEXShowcase] Error due to Recording/Setting parameters (#15)

Hi @RokasSthttps://github.com/RokasSt hope things are going well!
I guess you should also get the following warning at the start of the simulation:

Incompatible timeStep supplied: 50
Setting timeStep to 32 ms
Sample rate incompatible with timeStep
Sampling instead at 31 Hz

?
So the code that detects and manages incompatible time steps or sample rates assumes that your time step is always going to be <= 1, which is why it doesn't fix the time step you specify properly. Is there a reason you're using such a big time step of 50 ms (and such a slow sample rate)?

�
Reply to this email directly or view it on GitHubhttps://github.com//issues/15#issuecomment-137502822.

[richardtomsett]

OK no problem! I should change the VERTEX master to catch this and raise an error though I think, as people will get nonsensical results with long timesteps for the models VERTEX is meant to deal with @RokasSt